That metaphor is used in science is hardly a contentious issue. Extensive literature on the topic testifies to the widespread employment of metaphor in scientific texts. According to the received view, however, metaphor is a periphrastic 'as-if' device which has no bearing on the deliberative content of scientific theories. Metaphorical reasoning is taken to be epiphenomenal, a corollary to which is that metaphorical meaning is, in principle, reducible to equivalent literal meaning (for discussion, see section 5.2.1). Consequently, metaphors are considered ancillary in science and tend to be ascribed a marginal role as heuristic or exegetical tools to be used for pedagogical purposes. Contrary to this restrictive view on the function of metaphor in science, this study seeks to present the role of metaphor as both more significant and versatile. The aim of the study is to establish the mechanism of conceptual metaphor as operative in the constitution of scientific theories. Implicit in the thesis is the claim that metaphors are not merely linguistic expressions, but they are (primarily) a matter of reasoning. Thus, a prerequisite to an investigation of the role of metaphor in scientific theory articulation is demonstrating that metaphor operates at the level of conceptualisation. Only then can it be legitimately posited as a conceptual instrument of abstract reason. To this end, a framework of cognition is introduced in section IV in which metaphor is characterised naturally as a protective cognitive device whereby the cognitive agent establishes meaningful relationships between concepts and segments of external reality.